Page 3 of 5
Journal of the American Chemical Society
1
2
3
4
aYields are the average of two runs, 0.50 mmol scale. bOne run.
aAlkyl nucleophiles: RZnBr (3 equiv); Aryl nucleophiles: RZnBr
c0.20 mmol scale.
(1.5 equiv), ZnBr2 (1.5 equiv), DMA/THF; yields are the average
of two runs, 0.50 mmol scale. bOne run.
With optimized conditions in hand, we investigated the scope
of the aziridine coupling partner (Table 2). Whereas styrenyl
aziridines bearing para- and meta-substituents on the aromatic
ring can be alkylated in high yield (3, 9–14), those with orthoꢀ
substitution are not wellꢀtolerated (15 and 16), presumably due to
steric encumbrance in forming such quaternary centers. We were
pleased to find that a heteroaromatic aziridine underwent coupling
to afford 13 with high reaction efficiency. Furthermore, 1,1ꢀ
disubstituted aziridines bearing substituents other than methyl at
the benzylic position are also competent substrates, significantly
broadening the synthetic utility of the reaction (17–20). In general,
sensitive functional groups such as chlorides (9), amides (11),
silyl ethers (19) esters (20), sulfones (21), and anilines (22) are
wellꢀtolerated. However, 1,1ꢀ diarylaziridines fail to deliver the
coupled products under the same conditions or at elevated temꢀ
peratures. Finally, given the importance of sulfonamide moieties
in biologically relevant molecules,14 we decided to explore aziriꢀ
dines derived from the antihypertensive meticrane and the fluoꢀ
rescent dye dansyl amide. Gratifying, 21 and 22 can be obtained
under standard conditions in moderate but useful yields.
5
6
7
8
Notably, the products all feature the phenethylamine motif,
which comprises the core of biogenic amines.17 Despite their low
concentration in humans, these amines have been shown to play
key roles in the central nervous systems. Although many of their
derivatives show potential as pharmaceutical agents and biomediꢀ
cal probes, the lack of a robust and systematic method to prepare
analogs has been a hurdle for the evaluation of their activity.18 In
particular, those bearing a quaternary benzylic center have been
challenging to make and their utility has rarely been explored. The
current methodology provides an efficient route to these novel
structures.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Our previous studies on the Ni/EDO catalyst system for
Negishi couplings with monosubstituted aziridines revealed a
stereoablative mechanism for C–C bond formation.9a,b To evaluate
the stereochemical course of the present reaction, enantiopure (R)ꢀ
2j was subjected to the standard reaction conditions utilizing Froꢀ
DO as the ligand (eq 1). In accord with our previous reports,
significant racemization was observed in the formation of 17,
whereas recovered 2j remained enantiopure. This result is conꢀ
sistent with an irreversible oxidative addition, at or after which a
stereoablative step occurs. A likely mechanism for the oxidative
addition is singleꢀelectron transfer (SET) from a reduced nickel
species to generate a stabilized tertiary/benzylic radical intermediꢀ
ate. This mechanism presents the exciting opportunity to achieve
an asymmetric variant starting from racemic aziridines. Whereas
there are many robust methodologies for the racemic synthesis of
aziridines, asymmetric catalytic methods for their synthesis reꢀ
main quite limited.19 Thus, a stereoconvergent crossꢀcoupling
reaction would be compelling. Unfortunately, use of (+)ꢀFroꢀDO
for the coupling of (±)ꢀ2a with nꢀBuZnBr delivered racemic 3.
However, we were pleased to find that camphorsultam ligand (+)ꢀ
L4 delivered 3 in a promising 73% yield and 27% ee, demonstratꢀ
ing for the first time that an electronꢀdeficient olefin can serve as
chiral ligand for asymmetric alkyl cross coupling (eq 2).7 This
result also represents the first example of a stereoconvergent cross
coupling with a tertiary electrophile.
The nucleophile scope was subsequently studied (Table 3).
Alkyl zinc reagents bearing synthetically useful functional groups
such as acetals (24) chlorides (25), trifluoromethoxy groups (28),
as well as βꢀbranching (26) can be used in the coupling procedure.
Application of benzylzinc bromides as nucleophiles serves as a
rapid route to the synthesis of β,γꢀbisaryl amines from aziridines
(27–29). Although arylzinc reagents performed less efficiently in
our previous Negishi protocol with styrenyl aziridines,9a we found
that by modifying the preparation of these nucleophiles, in comꢀ
bination with using 1.5 equivalents of ZnBr2 as an additive,15
arylation of 1,1ꢀdisubstituted aziridines could be achieved with
improved yields (30–34). Paraꢀ and metaꢀsubstituted aryl zinc
reagents are suitable reaction partners, but orthoꢀsubstituted aryl
zinc reagents perform poorly (34). In addition, both electronꢀ
neutral and electronꢀpoor nucleophiles can be coupled in good
yield in contrast to what is possible under Friedel–Crafts condiꢀ
tions.16
Table 3. Scope of Organozinc Reagentsa
Our results raise the question: what key elements of the EDOs
in Table 1 are responsible for their significant differences in reacꢀ
tivity? One hypothesis for why ligand FroꢀDO imparts superior
efficiency to the other EDO ligands is that it features a more πꢀ
deficient olefin, which leads to a more favorable relative rate of
reductive elimination to βꢀhydride elimination from the Ni center.
However, examination of the 13C NMR shifts of the olefinic carꢀ
bons on these ligands reveals no correlation between the chemical
ACS Paragon Plus Environment